Pancreatic cancer remains one of the most deadly human malignancies. During the past decade, unprecedented progress has been made identifying the genetic basis for this disease, including the discovery of a number of common somatic mutations now confirmed to play important pathogenic roles. However, the recent application of whole genome deletion analysis and high throughput DNA sequencing has accelerated the rate of novel mutation detection well beyond any ability to functionally evaluate identified candidate genes. This mismatch between gene discovery and functional annotation will only increase with the completion of the already in-progress sequencing of the pancreatic cancer genome, an effort currently being pursued by investigators here at Johns Hopkins. In order to alleviate this bottleneck, and provide a system for higher throughput annotation of the pancreatic cancer genome, we have generated the first zebrafish model of exocrine pancreatic cancer. Based on the low costs and modest floorspace required to maintain adult zebrafish, as well as the ability to rapidly generate large numbers of transgenic lines, this organism offers many advantages in evaluating the molecular basis of human cancer. When an oncogenic version of human KRAS is expressed in developing zebrafish pancreas, pancreatic progenitor cells fail to undergo normal exocrine differentiation, leading to the subsequent formation of invasive pancreatic cancer. Zebrafish pancreatic cancers invade and metastasize, and exhibit many features in common with the human form of the disease, including abnormal activation of hedgehog signaling. In addition creating the first zebrafish model of exocrine pancreatic cancer, we have successfully generated transgenic lines in which a modified Gal4 transcriptional activator is expressed in pancreatic progenitor cells. Using transposon technology to insert UAS-regulated transgenes into the zebrafish genome, we now have the opportunity to functionally evaluate a wide variety of genetic lesions for their ability to modify pancreatic cancer initiation and/or progression, achieving a level of throughput not technically feasible in the mouse. Using these techniques, we now plan to pursue the following Specific Aims: First, to functionally annotate candidate dominant mutations identified in the pancreatic cancer genome, through their modular introduction into the zebrafish tumorigenesis model; second, to study the effects of graded changes in hMYC expression in pancreatic tumorigenesis, using an inducible Gal4/UAS system targeting progenitor cells in zebrafish exocrine pancreas; and third, to develop Cre-based models of KRAS-mediated pancreatic neoplasia in zebrafish. Together, these studies will provide important new information regarding the genetic basis for pancreatic cancer, allowing for the more rapid development of effective targeted therapies.